| In this work, we present a prototype model enabling us to study the mathematical behaviour of a computerized hierarchical questionnaire. Besides its hierarchical structure, the latter should have the following characteristics: (1) a closed set of predetermined acceptable answers is associated with every question, and (2) the questionnaire should return a certain numerical outcome based on the answers, each of these being associated with a numerical weight. The model uses a program called a results generator, which generates all possible results; thus the requirement to have a finite number of possible answers. The generator transforms the questionnaire into a mega-decision tree to process every possible result from which we derive a curve. From that curve (a depiction of a mathematical function), we propose a way to interpret the results. The model is applied to a real-life computerized hierarchical questionnaire to demonstrate its capabilities. The need to develop such a model arose from the creation of a software aiming to help a user to analyse the risk level of a SME (Small and Medium-sized Enterprise) project. The eRisC software was developed for and validated by entrepreneurs, economic agents, lenders and investors, to identify the main risk factors of SMEs development projects in order to improve their success rate and facilitate their financing. (Abstract shortened by UMI.)... |
